Sensor Fusion for Attitude Estimation and PID Control of ... · (b) Rotate about y-axis . Figure 5....
Transcript of Sensor Fusion for Attitude Estimation and PID Control of ... · (b) Rotate about y-axis . Figure 5....
Sensor Fusion for Attitude Estimation and PID
Control of Quadrotor UAV
Aminurrashid Noordin, Mohd Ariffanan Mohd Basri, and Zaharuddin Mohamed
Universiti Teknologi Malaysia, Skudai, Malaysia
Email: [email protected]; {ariffanan; zahar}@fke.utm.my
Abstract—This paper presents sensor fusion algorithm using
nonlinear complementary filter (NCF) for attitude
estimation and a proportional-integral-derivative (PID)
controller for small-scale quadrotor unmanned aerial
vehicle (UAV) stabilization. From inertial measurement unit
(IMU) data, gyroscope as a main sensor is fused to another
two sensors; accelerometer and magnetometer to correct
drift error of gyroscope to obtain reliable attitude
estimation. In this paper, the performance of NCF is
implemented on real-time while applying PID controller for
attitude stabilization of quadrotor UAV during hovering.
Experimental results show the effectiveness of PID
controller for quadrotor UAV stabilization during attitude
control.
Index Terms—nonlinear complementary filter, quadrotor,
PID, UAV.
I. INTRODUCTION
Microelectromechanical systems (MEMS) is a
combination of mechanical and electrical components
into microscale objects. In advanced of MEMS
technology, components being small size, less expensive,
and low power consumption, sensors such as inertial
measurement unit (IMU) are redesigned to include onto
devices such as smartphone, automotive industries and
others application.[1]-[4].
In robotics, IMU is the main of attitude and heading
reference system (AHRS) to determine rotation, motion,
location and direction (generally called attitude
estimation) of mobile robots for automated navigation,
autonomous underwater vehicle (AUV) for global
localization systems and unmanned aerial vehicle (UAV)
in aviation [1], [4]-[6]. Hence, research community and
DIY hobbyist take this opportunity, utilizes small scale
IMU and focus on development of unmanned aerial
vehicles (UAV) which is very promising vehicle for
navigations, surveillances, and as well as educational
purposes [7]-[10].
However, IMU sensor performance is commonly
effected by biases and noises which tend to drift over
time for especially gyroscope and reduce the accuracy of
measurement. Therefore, most researcher applied sensor
fusion algorithms techniques to overcome the
measurement errors and obtaining accurate reading [1],
[4], [5], [11].
Manuscript received January 5, 2018; revised July 10, 2018.
Quadrotor UAV is an aerial vehicle that has
capabilities in vertical take-off and landing (VTOL),
omni-directional flying, and easy hovering performances
in limited spaces and always being considered in research
due to the simplest electronics and mechanical structures
design. However, quadrotor UAV is an under-actuated
and dynamically unstable system which possess with
complex behaviours. Many presented work in literatures
use ‘+’ configuration and simplified model, where non-
linear effect is neglected. Several literatures have
mentioned of proportional-integral-derivative (PID)
control a quadrotor [12]-[15] but using linearize model.
Thispaper focus on low cost IMU fusion using
nonlinear complementary filter for attitude estimation of
quadrotor UAV and PID controller for highly nonlinear
quadrotor UAV. The system comprises of quadrotor F450
frame (x configuration model), and APM2.6 flight
controller with built in IMU (MPU6000) and external
HMC5833L compass sensor. Data from low frequencies
parts of accelerometer and magnetometer are fused to the
high frequency part of the gyroscope signal. Drift error
by the main source (gyroscope) is corrected by
accelerometer and magnetometer. A PID controller is
used for attitude stabilization during quadrotor hovering.
A flipped test is conducted to observed quadrotor
performance during initial start on a rotating test-bed
fixed on one axis to ensure its stabilization before free
test on 3dof experimental platform with 𝑘𝑝, 𝑘𝑖 , and 𝑘𝑑
parameter setting.
This paper is structured as follows. Section II briefly
described the nonlinear quadrotor UAV modelling in
Newton-Euler formulation. Section III explained
regarding sensors used in IMU such as gyroscope,
accelerometer and magnetometer. Section IV mentioned
sensor fusion algorithm technique used in this research.
Section V shows the PID controller design method for
attitude stabilization of quadrotor UAV. Section VI gives
the experiments results and conclusions are stated in
Section VII.
II. QUADROTOR MODEL
Quadrotor UAV is a type of helicopter that can be
controlled by varying the rotor speeds. It is an under-
actuated, dynamic vehicle with four input forces and six
output coordinates. Quadrotor UAV composed of four
rotors with symmetrically arrangement where two
International Journal of Electrical and Electronic Engineering & Telecommunications Vol. 7, No. 4, October 2018
©2018 Int. J. Elec. & Elecn. Eng. & Telcomm. 183doi: 10.18178/ijeetc.7.4.183-189
diagonal motors (1 and 2) are running in the same
direction whereas the others (3 and 4) in the other
direction to eliminate the anti-torque [10], [14], [16].
Quadrotor UAV have been designed with
symmetrically structure in either ‘x’ mode configuration
or ‘+’ mode configuration. This research used ‘x’ mode
configuration quadrotor as illustrated in Fig. 1, where the
coordinate systems of two reference frames describe the
dynamics of a quadrotor; an earth fixed initial reference
frame, {E} and a body fixed reference frame {Q} located
at the center of gravity (COG) of quadrotor body frame
which is a rigid body in free motion with six Degree of
Freedom (DOF) consist of three translational and three
rotational.
Figure 1. Inertial coordinate systems and body-fixed frame for x
configuration quadrotor
For the modelling, the following assumption is
defined for simplification [17]:
1. The quadcopter is assumed as a rigid body.
2. The quadcopter’s structure is assumed as
symmetric with respect to the XY-axis.
3. The centre of mass and the origin of the body
fixed frame are coinciding.
4. The propellers are considered as rigid; no blade
flapping occurs.
5. The four propellers work under the same
conditions at any time, meaning that thrust
coefficient, and reaction torque coefficient, are the
same for all propellers.
The generalized coordinates for the quadrotor based
on Fig. 1 can be described as follow:
𝜉 = [𝑥, 𝑦, 𝑧]𝑇 ∈ ℝ3, 𝜂 = [𝜙, 𝜃, 𝜓]𝑇 ∈ ℝ3 (1)
whereas, in (1) vector 𝜉 , denotes the position of the
quadrotor relative to inertial frame, vector 𝜂, denotes the
attitude of the quadrotor. The relation of body fixed
reference frame, {Q} respect to earth fixed initial
reference frame, {E} satisfy as {𝑄}𝑇 = 𝑅𝑇 × {𝐸}𝑇 .
Equation (2) defines the rotation matrix 𝑅𝑇, where, S and
C stands for trigonometric operators ‘sin’ and ‘cos’
respectively.
𝑅𝑇 = [
𝐶𝜓𝐶𝜃 −𝑆𝜓𝐶𝜙 + 𝐶𝜓𝑆𝜃𝑆𝜙 𝑆𝜓𝑆𝜙 + 𝐶𝜓𝐶𝜙𝑆𝜃𝑆𝜓𝐶𝜃 𝐶𝜓𝐶𝜙 + 𝑆𝜓𝑆𝜃𝑆𝜙 −𝐶𝜓𝑆𝜙 + 𝐶𝜙𝑆𝜓𝑆𝜃−𝑆𝜃 𝐶𝜃𝑆𝜙 𝐶𝜃𝐶𝜙
] (2)
From general Newton-Euler translational and
rotational dynamics, the quadrotor dynamics, is described
as (3) and (4) respectively, where, 𝑔 is gravitational
coefficient, 𝐸𝑧 is vector matrix of z-axis defined as
[0 0 1]𝑇 , 𝑈1 is total thrust force generated by four rotors.
𝐼 is the moments of inertia for the quadrotor, a diagonal
matrix 3-by-3 and defined as 𝐼 = diagonal[𝐼𝑥𝑥𝐼𝑦𝑦𝐼𝑧𝑧]𝑇. 𝐽𝑟
is rotor inertia, Ω𝑑is total rotor speeds generated from the
two pairs of rotor. 𝑈2, 𝑈3 and 𝑈4are total torque, 𝜏related
to quadrotor as of total summation of Coriolis torque, 𝜏𝑐,
and Gyroscopic torque, 𝜏𝑔 and quadrotor body frame
torque, 𝜏𝑎.
𝑚𝜉̈ = 𝑚𝑔𝐸𝑧 + 𝑈1𝑅𝑇𝐸𝑧 (3)
𝐼�̈� = −𝜂 × 𝐼𝜂 − 𝐽𝑟(𝜂 × 𝐸𝑧)Ω𝑑 + [𝑈2𝑈3𝑈4]𝑇 (4)
Finally, from (3) and (4), the final equation for
quadrotor translation dynamics and rotational dynamics
can be formulated as
𝑚�̈� = 𝑈1(𝑆𝜓𝑆𝜙 + 𝐶𝜓𝑆𝜃𝐶𝜙)
𝑚�̈� = 𝑈1(−𝐶𝜓𝑆𝜙 + 𝑆𝜓𝑆𝜃𝐶𝜙)
𝑚�̈� = 𝑚𝑔 + 𝑈1(𝐶𝜃𝐶𝜙) (5)
𝐼𝑥𝑥�̈� = (𝐼𝑦𝑦 − 𝐼𝑧𝑧)�̇��̇� − (𝐽𝑟Ω𝑑)�̇� + 𝑙𝑈2
𝐼𝑦𝑦�̈� = (𝐼𝑧𝑧 − 𝐼𝑥𝑥)�̇��̇� + (𝐽𝑟Ω𝑑)�̇� + 𝑙𝑈3
𝐼𝑧𝑧�̈� = (𝐼𝑥𝑥 − 𝐼𝑦𝑦)�̇��̇� + 𝑈4 (6)
III. INERTIAL MEASUREMENT UNIT
With advanced in MEMS, 3-axis gyroscopes are often
implemented with a 3-axis accelerometer to provide a full
6 degree-of-freedom (DOF) motion tracking system for
many applications such unmanned aerial vehicles (UAV),
mobile robot and smartphone. With 3-axis digital
compass additional of 3-DOF can assist on a heading for
a system for correct orientation during locomotion. Fig. 2
shows 6-DOF MPU6000 comprise of gyro and
accelerometer. 3-axis compass can be attached as
auxiliary sensor for attitude estimator system such used in
APM2.6.
Figure 2. A 6-DOF – Gyro/Accelerometer
A. Gyroscope
A gyroscope is a device typically used for navigation
by measurement of angular velocity and estimated
changing in orientation [18]. However, for a long term,
Gyroscope reading tend to drift from static condition.
Gyroscopes can measure rotational velocity up to three
directions. In IMU, a gyroscope or gyros sensor as shown
in Fig. 3 is used to detect rotation speed to measure
angular velocity or in other term called angular rate with
measurement unit degree per second (°/s) or revolution
per second (RPS).
International Journal of Electrical and Electronic Engineering & Telecommunications Vol. 7, No. 4, October 2018
©2018 Int. J. Elec. & Elecn. Eng. & Telcomm. 184
Figure 3. A 3-Axis Gyro Sensor (Source: Adafruit Industries, 2017)
By integrating the angular rate (reading from sensor)
of each axis, the roll angle (𝜙), the pitch angle (𝜃), and
the yaw angle (𝜓) can be obtained using (7) as follow
𝜙 = ∫ 𝜔𝑥 𝑑𝑡
𝜃 = ∫ 𝜔𝑦 𝑑𝑡
𝜓 = ∫ 𝜔𝑧 𝑑𝑡 (7)
Standard control range of angle in aviation system is
between −180 to 180 degrees. Therefore, result from (7)
need to be adjusted within this range.
B. Accelerometer
An accelerometer sensor as shown in Fig. 4 is a
dynamic MEMs sensor used to measure acceleration
forces up to three orthogonal axes. Measurement unit of
accelerometer typically in gravitational motion (g) where
1g is equivalent to 9.8m/s2 which is acceleration cause of
earth gravity. However, for linear displacement
measurement, recorded data need to be integrated two
times which tend to drift over a time and very sensitive to
environmental noise [3], [19].
In other way, from an accelerometer, a tilt angle can be
sense and estimate as described on Freescale
Semiconductor Application Note used by [20].
Figure 4. A 3-Axis Accelerometer Sensor (Source: Adafruit Industries, 2017)
(a) Rotate about x-axis
(b) Rotate about y-axis
Figure 5. Three Axes tilt sensing for (a) Roll angle and (b) Pitch angle
For three axes rotation, the same concept is applied as
shown in Fig. 5 where roll angle (𝜙) and pitch angle (𝜃)
can be computed as (8) and (9) respectively. The rotation
follows typical right-hand rules law:
𝜙 = tan−1(𝐴𝑦/√𝐴𝑥2 + 𝐴𝑧
2) (8)
𝜃 = tan−1(𝐴𝑥/√𝐴𝑦2 + 𝐴𝑧
2) (9)
C. Magnetometer
A magnetometer as shown in Fig. 6 can sense where
the strongest magnetic force is coming from, generally
used to detect magnetic north. Therefore, as for digital
compass, magnetometer sensor is used to provide heading
information where the heading system calculation
depending to magnetic data (X, Y, Z) therefore the
compass orientation can be mathematically rotated to
horizontal plane as show in Fig. 7. Based on this figure, X,
Y, and Z from magnetic sensor reading can be
transformed to the horizontal plane (Xh, Yh) by applying
the rotation as (10) with aided from tilt sensing roll (𝜙)
and pitch (𝜃) from an accelerometer sensor. As such it is
useful to determine absolute orientation in the NEWS
plane of any systems during navigation [20].
𝑋ℎ = 𝑚𝑥 cos(𝜙) + 𝑚𝑦 sin(𝜙) − 𝑚𝑧 cos(𝜃) sin(𝜙)
𝑌ℎ = 𝑚𝑦 cos(𝜃) + 𝑚𝑧sin (𝜃) (10)
From (10) then, the heading information can be
calculated as
Heading = 𝜓ℎ = atan2 (𝑌ℎ
𝑋ℎ) (11)
Figure 6. A 3-Axis Magnetometer Sensor (Source: Adafruit Industries, 2017)
Figure 7. Tilt sensing angle in local horizontal plane defined by gravity
(Caruso, 2000)
IV. SENSOR FUSION ALGORITHM
Based on [20] and [21], in this sensor fusion,
gyroscope signal is used as references for orientation of
the UAV to update the direct cosine matrix (DCM).
Accelerometer and magnetometer are used to correct drift
error from gyroscope and heading error. A feedback
proportional integral (PI) controller is used to eliminate
the steady state error in this system.
International Journal of Electrical and Electronic Engineering & Telecommunications Vol. 7, No. 4, October 2018
©2018 Int. J. Elec. & Elecn. Eng. & Telcomm. 185
DCM for rigid body frame respect to earth frame
expressed in Euler angles is described as (3). For small
angle rotation, 𝑆𝜙 → 𝜙, 𝑆𝜃 → 𝜃, 𝑆𝜓 → 𝜓 and the cosines
of these angles approach unity. By neglecting the
products of the angles because become small, the DCM
can be simplified to the skew symmetric form shown as
below:
𝑅𝐵𝐸 ≈ [
1 −𝜓 𝜃𝜓 1 −𝜙
−𝜃 𝜙 1] (12)
In kinematics, the rate of change in rotating vector
gives by
𝑑𝑟(𝑡)
𝑑𝑡= 𝑤(𝑡) × 𝑟(𝑡) (13)
where 𝑤(𝑡) is angular rate vector and r(t) is rotating
vector with same reference frame. By assuming signals of
gyroscope not affected by drift, the matrix to update
DCM from gyroscope signals can be described as
𝑅(𝑡 + 𝑑t) ≈ 𝑅(𝑡) [
1 −𝑑𝜓 𝑑𝜃𝑑𝜓 1 −𝑑𝜙
−𝑑𝜃 𝑑𝜙 1] (14)
To maintain the orthogonality of rotating frame,
renormalize is needed. The first step is to obtain the error
that measures the rotation of X and Y toward each other
by dot product
err = 𝑋 ∙ 𝑌 = 𝑋𝑇𝑌 = [𝑟𝑥𝑥 𝑟𝑥𝑦 𝑟𝑥𝑧] [
𝑟𝑦𝑥
𝑟𝑦𝑦
𝑟𝑦𝑧
] (15)
Then the orthogonal value of X, Y and Z can be
computed as
𝑋𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙 = 𝑋 −𝑒𝑟𝑟𝑜𝑟
2𝑌
𝑌𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙 = 𝑌 −𝑒𝑟𝑟𝑜𝑟
2𝑋
𝑍𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙 = 𝑋𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙 × 𝑌𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙 (16)
The final step is to ensure magnitude of each row of R
matrix is equal to one. This is done by scaling the row of
the R matrix using Taylor’s expansion as
𝑋𝑛𝑜𝑟𝑚 =1
2(3 − 𝑋𝑜𝑟ℎ𝑡𝑜𝑔𝑜𝑛𝑎𝑙 ∙ 𝑋𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙)𝑋𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙
𝑌𝑛𝑜𝑟𝑚 =1
2(3 − 𝑌𝑜𝑟ℎ𝑡𝑜𝑔𝑜𝑛𝑎𝑙 ∙ 𝑌𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙)𝑌𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙
𝑍𝑛𝑜𝑟𝑚 =1
2(3 − 𝑍𝑜𝑟ℎ𝑡𝑜𝑔𝑜𝑛𝑎𝑙 ∙ 𝑍𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙)𝑍𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙 (17)
The roll-pitch error is computed by taking the cross
product of the reference measurement of gravity with the
Z row of DCM and can be described by
𝑒𝑟𝑜𝑙𝑙−𝑝𝑖𝑡𝑐ℎ = [𝑟𝑧𝑥 𝑟𝑧𝑦 𝑟𝑧𝑧]𝑇
× 𝑔𝑟𝑒𝑓 (18)
The yaw error in the earth frame is the z component of
the cross product of the Heading vector and the x column
of the R matrix
𝑒𝑦𝑎𝑤(𝑒𝑎𝑟𝑡ℎ) = 𝑟𝑥𝑥𝐻𝑒𝑎𝑑𝑖𝑛𝑔𝑦 − 𝑟𝑦𝑦𝐻𝑒𝑎𝑑𝑖𝑛𝑔𝑥 (19)
Then, the yaw error in body frame can be computed as
𝑒𝑦𝑎𝑤 = 𝑒𝑦𝑎𝑤(𝑒𝑎𝑟𝑡ℎ) [
𝑟𝑧𝑥
𝑟𝑧𝑦
𝑟𝑧𝑧
] (20)
Finally, the total error vector can be written as
𝑒𝑡𝑜𝑡𝑎𝑙 = 𝑊𝑟𝑝𝑒𝑟𝑜𝑙𝑙−𝑝𝑖𝑡𝑐ℎ + 𝑊𝑌𝑒𝑦𝑎𝑤 (21)
V. PID CONTROL
Proportional Integral Derivative (PID) controller is a
classical controller that have proven to be robust and
tremendously beneficial in many linear or non-linear
applications. The PID design are pointed out in many
references, such as [13], [22]-[25]. The controller
attempts to minimize the error over time by adjustment of
a control variable 𝑢(𝑡). The mathematical representation
of PID controller is given as
𝑢(𝑡) = 𝑘𝑝𝑒(𝑡) + 𝑘𝑖 ∫ 𝑒(𝑡) + 𝑘𝑑𝑑
𝑑𝑡𝑒(𝑡) (22)
where, 𝑢(𝑡) is the input signal and the error signal 𝑒(𝑡) is
defined as
𝑒(𝑡) = 𝑑𝑒𝑠𝑖𝑟𝑒𝑑_𝑖𝑛𝑝𝑢𝑡(𝑡) − 𝑎𝑐𝑡𝑢𝑎𝑙_𝑜𝑢𝑡𝑝𝑢𝑡(𝑡) (23)
On the other hand, a PID controller continuously
calculates an error value 𝑒(𝑡) then applies a correction
based on proportional, integral, and derivative terms as
shown in Fig. 8.
The proposed controllers for attitude and altitude
stabilization in this experiment are shown as Fig. 9,
where a cascade PID is used. For attitude stabilization in
the inner loop PID controller, angular rate �̇� , �̇� , �̇� and
velocity of �̇� are used as references input and have its
own gain. For altitude control in outer loop PID
controller, desired 𝑧 , 𝜙 , 𝜃 , and 𝜓 is set as desired
input/reference.
Figure 8. PID controller structure
Figure 9. Cascade PID in nonlinear quadrotor’s attitude stabilization and altitude control
International Journal of Electrical and Electronic Engineering & Telecommunications Vol. 7, No. 4, October 2018
©2018 Int. J. Elec. & Elecn. Eng. & Telcomm. 186
Figure 10. Avionics system for quadrotor
a) Initial condition b) On going experiment
Figure 11. Single axis platform for flipping
TABLE I. PID PARAMETERS FOR QUADROTOR UAV
Angles Kp Ki Kd
Stabilize
𝝓 5 0.003 0
𝜽 5 0.003 0
𝝍 7 0.02 0
Rate
�̇� 0.2 0.004 0.003
�̇� 0.2 0.004 0.003
�̇� 0.13 0.02 0
VI. EXPERIMENTAL RESULTS
Fig. 10 shows flight control system for the quadrotor
that used in this experiment consist of APM2.6 (based on
Arduino MEGA) with built in IMU (accelerometer and
gyroscope) and external compass with GPS module.
Several experiments have been conducted to test
quadrotor UAV stabilization while tuning the PID
parameters. For the first test, the quadrotor UAV is tied
up on a single road testbed platform. Fig. 11 (a) and Fig.
11 (b) shows the quadrotor UAV condition tied up on the
testbed before and during the experiment. In this
experiment, the quadrotor UAV is only allowed to rotate
on roll, 𝜙 angle. The PID parameters is tuned until the
quadrotor UAV able to flip 180° and stabilized at 0°. The
chosen PID parameter is stated on Table I.
Fig. 12 shows roll angle performance during flipping
test. Initially, the quadrotor UAV is at 180° position and
after the throttle is set about 50% of maximum speed, it
immediately rotates to 0° in 1 to 2 seconds and try to be
stabilized until at 13 seconds before the throttle is set to
zero. Since quadrotor UAV still has momentum, it still
rolling for a few second before stabilized until 26 seconds
before the throttle once again is set about 50% of
maximum speed. The quadrotor UAV once again able to
be flipped from 180° to 0° and stabilized. It shows that
with the chosen PID parameter, the quadrotor UAV able
to maintain it stability during hovering.
Then the experiment continues with 3dof testbed
platform as shows in Fig. 13. Here, the quadrotor UAV
being test to response on roll, pitch and yaw angle input
from radio controller.
Figure 12. Roll, 𝜙 angle performance during flipping test
Figure 13. 3dof experimental platform
Fig. 14, Fig. 15 and Fig. 16 show roll, pitch and yaw
response to the input from radio controller respectively.
During this experiment, the throttle is set about 50% of
maximum input. The roll and pitch from radio controller
are push right and left, upward and backward respectively
to see quadrotor UAV response for stabilization during
hovering.
Telemetry
Radio controller
Telemetry
Ground station APM2.6 Quadrotor Compass &
GPS module
International Journal of Electrical and Electronic Engineering & Telecommunications Vol. 7, No. 4, October 2018
©2018 Int. J. Elec. & Elecn. Eng. & Telcomm. 187
Figure 14. Roll angle response
Figure 15. Pitch angle response
Figure 16. Yaw angle response
As shown in Fig. 14, at 30 seconds, the quadrotor
UAV try to stabilize at zero (0) degree after being push to
right and left. The positive and negative spikes show the
quadrotor UAV angle response to the radio controller
input from roll. Similarly, as shown in Fig. 15, where at
50 seconds, the quadrotor UAV try to stabilize at zero (0)
degree after being upward and backward. The positive
and negative spikes show the quadrotor UAV angle
response to the radio controller input from pitch.
Fig. 16 shows a response from yaw direction where the
quadrotor UAV able to rotate right and left between zero
(0) to 360-degree angle response to the radio controller
input from yaw.
VII. CONCLUSIONS
This paper has presented a sensor fusion based on
nonlinear complementary filter algorithm and PID
controller design for the quadrotor UAV for attitude
estimation and stabilization during hovering. The sensor
fusion algorithm used gyroscope data (high frequency
part) as the main source and the drift corrected by
accelerometer and magnetometer data (low frequencies
part).The proposed PID controller is designed based on
the cascade control approach where, for attitude
stabilization in the inner loop PID controller, angular rate
�̇�, �̇�, �̇� and velocity of �̇� are used as references input and
have its own gain. For altitude control in outer loop PID
controller, desired 𝑧 , 𝜙 , 𝜃 , and 𝜓 is set as desired
input/reference. Two experiments have been conducted
for flipped and attitude stabilization. PID parameters is
tuned to ensure the quadrotor UAV able to flip and
stabilize from 180-degree initial position for roll axis.
Then the quadrotor was tested on 3dof experimental
platform for attitude stabilization respond to input from
radio controller. Experimental results demonstrated the
effectiveness of the proposed approach throughout the
test.
Further research needs to be done to extend the results
shown in the present paper with input disturbance such as
wind while outdoor flight test. Different type of controller
can be further investigated to improve system
performances.
ACKNOWLEDGMENT
The authors would like to thank Universiti Teknologi
Malaysia (UTM) under the Fundamental Research Grant
Scheme (R.J130000.7823.4F761), Research University
Grant (Q.J130000.2523.15H39), Universiti Teknikal
Malaysia Melaka (UTeM), and Ministry of Higher
Education for supporting this research.
REFERENCES
[1] Y. Hu, Y. Yan, J. Liang, and L. Wang, “A miniature, low-cost
MEMS AHRS with application to posture control of robotic fish,” in Proc. IEEE Instrum. Meas. Technol. Conf., 2013, pp. 1392–
1395.
[2] B. Allotta, R. Costanzi, and F. Fanelli, “An attitude estimation algorithm for underwater mobile robots,” IEEE/ASME Trans.
Mechatronics, vol. 21, no. 4, pp. 1900–1911, 2016. [3] U. Guner, H. Canbolat, and A. Unluturk, “Design and
implementation of adaptive vibration filter for MEMS based low
cost IMU,” in Proc. 9th Int. Conf. Electr. Electron. Eng., 2015, pp. 130–134.
[4] L. Wang, L. Fu, X. Hu, and G. Zhang, “Attitude estimation for UAV with low-cost IMU/ADS based on adaptive-gain
complementary filter,” in Proc. Int. Symposium on Neural
Networks, 2016, pp. 346-355. [5] Z. Wu, Z. Sun, W. Zhang, and Q. Chen, “Attitude and gyro bias
estimation by the rotation of an inertial measurement unit,” Meas. Sci. Technol., vol. 26, no. 12, p. 125102, 2015.
[6] D. Q. Duong, J. Sun, T. P. Nguyen, and L. Luo, “Attitude
estimation by using MEMS IMU with fuzzy tuned complementary
International Journal of Electrical and Electronic Engineering & Telecommunications Vol. 7, No. 4, October 2018
©2018 Int. J. Elec. & Elecn. Eng. & Telcomm. 188
filter,” in Proc. IEEE Int. Conf. Electron. Inf. Commun. Technol., 2016, pp. 372–378.
[7] Z. R. Mahayuddin, H. M. Jais, and H. Arshad, “Comparison of
human pilot (remote) control systems in multirotor unmanned aerial vehicle navigation,” Int. J. Adv. Sci. Eng. Inf. Technol., vol.
7, no. 1, pp. 132–138, 2017. [8] J. Ajmera and V. Sankaranarayanan, “Point-to-point control of a
quadrotor: Theory and experiment,” IFAC-PapersOnLine, vol. 49,
no. 1, pp. 401–406, 2016. [9] J. Bazin, T. Fields, and A. J. Smith, “Feasibility of in-flight
quadrotor individual motor thrust measurements,” in Proc. AIAA Atmospheric Flight Mechanics Conf., 2016, pp. 1–12.
[10] B. B. V. L. Deepak and P. Singh, “A survey on design and
development of an unmanned aerial vehicle (quadcopter),” Int. J. Intell. Unmanned Syst., vol. 4, no. 2, pp. 70–106, 2016.
[11] Y. Wang, N. Li, X. Chen, and M. Liu, “Design and implementation of an AHRS based on MEMS sensors and
complementary filtering,” Adv. Mech. Eng., vol. 2014, 2014.
[12] R. García, F. R. Rubio, and M. G. Ortega, “Robust PID control of the quadrotor helicopter,” IFAC Proc. Volumes, vol. 45, no. 3, pp.
229-234, 2012. [13] F. Goodarzi, D. Lee, and T. Lee, “Geometric nonlinear PID
control of a quadrotor UAV on SE (3),” in Proc. Eur. Control
Conf., 2013, pp. 3845–3850. [14] J. Ghommam, L. F. Luque-Vega, B. Castillo-Toledo, and M. Saad,
“Three-dimensional distributed tracking control for multiple quadrotor helicopters,” J. Franklin Inst., vol. 353, no. 10, pp.
2344–2372, 2016.
[15] S. Bouabdallah, A. Noth, R. Siegwart, and R. Siegwan, “PID vs LQ Control Techniques Applied to an Indoor Micro Quadrotor,”
in Proc. IEEE/RSJ Int. Conf. Intell. Robot. Syst., 2004, pp. 2451–2456.
[16] D. Domingos, G. Camargo, and F. Gomide, “Autonomous fuzzy
control and navigation of quadcopters,” IFAC-PapersOnLine, vol. 49, no. 5, pp. 73–78, 2016.
[17] M. A. M. Basri, A. R. Husain, and K. A. Danapalasingam, “Enhanced backstepping controller design with application to
autonomous quadrotor unmanned aerial vehicle,” J. Intell. Robot.
Syst. Theory Appl., vol. 79, no. 2, pp. 295–321, 2014. [18] H. Fourati, N. Manamanni, L. Afilal, and Y. Handrich,
“Complementary observer for body segments motion capturing by inertial and magnetic sensors,” IEEE/ASME Trans. Mechatronics,
vol. 19, no. 1, pp. 149–157, 2014.
[19] N. H. Ariffin, N. Arsad, and B. Bais, “Low cost MEMS gyroscope and accelerometer implementation without Kalman Filter for angle
estimation,” in Proc. Int. Conf. Adv. Electr. Electron. Syst. Eng., 2016, pp. 77–82.
[20] D. Derawi and J. Kim, “Real-time nonlinear complementary observer for low-cost inertial attitude system,” presented at 2013
Int. Glob. Navig. Satell. Syst. Soc. IGNSS Symp., 2013.
[21] M. S. Saelal, D. Derawi, N. D. Salim, and M. Z. M. Tumari, “Real-time nonlinear complementary filter on SO(3) for attitude
estimation of small-scale ariel robot,” in Proc. Int. Conf. Vision, Image Signal Process., 2017, pp. 109–113.
[22] J. Li and Y. Li, “Dynamic analysis and PID control for a
quadrotor,” in Proc. IEEE Int. Conf. Mechatronics Autom., 2011, pp. 573–578.
[23] S. Liu, Z. Hou, and J. Zheng, “Attitude adjustment of quadrotor aircraft platform via a data-driven model free adaptive control
cascaded with intelligent PID,” in Proc. 28th Chinese Control
Decis. Conf., 2016, pp. 4971–4976. [24] B. Kada, “A new methodology to design sliding-PID controllers:
Application to missile flight control system,” presented at IFAC Conf. on Advances in PID Control, 2012.
[25] M. Nguyen Duc, T. N. Trong, and Y. S. Xuan, “The quadrotor
MAV system using PID control,” in Proc. IEEE Int. Conf. Mechatronics Autom., 2015, pp. 506–510.
Aminurrashid Noordin received the B.Eng. and the M.Eng. Degree in
Mechatronics Engineering from Universiti Teknologi Malaysia in 2002
and 2009 respectively, where he is currently working toward the Ph.D. in the Department of Control and Mechatronics Engineering of
Universiti Teknologi Malaysia (UTM). Since 2011, he has been with Department of Electrical Engineering Technology, Faculty of
Engineering Technology of Universiti Teknikal Malaysia Melaka where
he is currently a Lecturer. His research interests include Nonlinear Control System, Robotics and Embedded System.
Mohd Ariffanan Mod Basri received the B.Eng. and the M.Eng.
Degree in Mechatronics Engineering from Universiti Teknologi
Malaysia in 2004 and 2009 respectively. He also received the Ph.D. in Electrical Engineering from Universiti Teknologi Malaysia in 2015. He
is currently a Senior Lecturer in Department of Control and Mechatronics Engineering of Universiti Teknologi Malaysia. His
research interests include intelligent and Nonlinear Control Systems.
Zaharuddin Mohamed received his B.Eng in Electrical, Electronics
and Systems from Universiti Kebangsaan Malaysia (UKM) in 1993, M.Sc. in Control Systems Engineering from The University of Sheffield
in 1995 and Ph.D. in Control Systems Engineering from The University
of Sheffield in 2003. Currently, he is an Associate Professor in Department of Control and Mechatronics Engineering of Universiti
Teknologi Malaysia (UTM) and his current research interest involve the control of Mechatronics systems, flexible and smart structures.
International Journal of Electrical and Electronic Engineering & Telecommunications Vol. 7, No. 4, October 2018
©2018 Int. J. Elec. & Elecn. Eng. & Telcomm. 189